The present invention relates to a system support technique for facilitating change of a database system to be implemented using a plurality of middleware bases such as data warehouses (DWH). More particularly, the present invention is directed to modifying a modification location in a plurality of middleware bases and to a modifying the specified location. The present invention also relates to a system modification support of operations of forming such database information as not only data warehouse but also a web system to be disclosed on the Web.
In these years, in a wide range of various types of shops or stores such as department stores, specialty stores, mass merchandisers hawking their electronic wares or domestic appliances, and super markets, the attribute data (such as name and address) of a customer who bought articles and his or her purchase history can be acquired using his or her unique payment card which can be used between shops or stores or between store or shop groups. In such circumstances, there has been formed a system which processes purchase histories acquired in individual shops or stores according to analysis techniques having various purposes. In such background, there is a system called a data warehouse which collects detail information about purchase histories or the like from various viewpoints and performs detail searching operation.
A data warehouse is regarded as belonging to one of database system fields. Since the data warehouse outputs data results associated with various viewpoints, the data warehouse is implemented, in many cases, with use of many middleware bases including relational database (RDB) and multidimensional database (MDB), an extraction, transformation and loading (ETL) tool, and a schedule management tool of data processing job.
The multidimensional database is a database which is different from a related relational database and has three or more dimensions. In the relational database, data structure for data storage is called a table; while, in the multidimensional database, a data structure for data storage is called a cube. For example, as an example of the cube, an age/department/period cube 209 having an age dimension, a department dimension and a period dimension is shown in FIG. 2. The multidimensional database can operate to display slicing data, raise or lower the dimensional layer which is an object to be analyzed.
The multidimensional database is described in William A. Giovinazzo, “Object-Oriented Data Warehouse Design Building a Star Schema”, pp. 1–2, pp. 127–152, PRehTICE HALL, 2000.
In this connection, a tendency is to strongly demand of shortening the period of forming an information system such as the above data warehouse. With respect to the data warehouse, for the purpose of shortening the period of forming a system in response to customer's demand, there are known techniques which follow.
As a first related art, “StartNow for Relationship Marketing” (which is a registered trademark of IBM Corporation) of IBM Corporation as packages grouped by different business types or businesses integrally offers hardware, software, analysis model and introduction/support services grouped by business types, which are specialized in customer relationship management (CRM).
As a second related art, “RedBrick Warehouse” of Infocom company as a tool for supporting formation or construction of a data warehouse is a database exclusive to the construction of the data warehouse. Further, a tool called “Red Brick Warehouse Administrator” for supporting the design and construction of a data warehouse offers a graphic user interface (GUI) for defining a relationship between RDB tables forming the data warehouse and a graphic user interface (GUI) for calculating an approximate value of a necessary disk capacity. The “RedBrick” is described in Robert J. Hocutt, “The official guide to Informix/Red Brick Data Warehousing”, chapter 1, Hurgry Minds, Inc., 2000 (“RedBrick Warehouse” being a registered trademark of Infocom company).
Oracle Corporation also supplies “Oracle Warehouse Builder (OWB)” for designing, mounting and managing a data warehouse. Using the OWB, the RDB table, MD schema and ETL job can be designed through GUI. (“Oracle Warehouse Builder (OWB)” is a registered trademark of Oracle Corporation).
The first related art is directed to shortening the introduction period using the package as a target. However, since the data warehouse and analysis system are systems for supporting strategy-making or decision-making based on features of its own firm or business model, the system is required to be altered or modified by an introduction destination user according to the circumstances of the introduction destination. However, the internal structure of the package is implemented on a plurality of middleware bases, it is difficult to identify a modification location or to perform modifying operation the location according to the above alternation. The modification location is also limited to within a range assumed as the package. For this reason, when a demand of such alternation takes place, it sometimes becomes impossible to shorten the construction period even when such a package as described in the related art 1 is used. Further, when it is tried to cope with various types of customer requests using such a related art, it is required to manage many parts to cope with the alternation of pre-estimated modifications, which leads to the fact that the number of parts managed by the system is increased and system management becomes complicated.
In particular, the formation or construction of the data warehouse is implemented with use of many middleware bases including RDB, MDB, and a job management system for managing data processing jobs on the RDB or MDB. As a result, even when the alternation is small in external specifications, it is required to alter the operation on the plurality of middleware bases. For this reason, (1) since the modification location is dispersed on a plurality of middleware bases, the number of steps necessary for detecting the modification location is increased, (2) since the detected modification location is modified in the developing environment of the plurality of middlewares, the number of steps necessary for maintaining the consistency or compatibility or for confirming operations becomes large, and (3) since the developer is required to have a high level of knowledge, the modifying operation cannot be carried out without such an operator having such a high level of knowledge.
The second related art is directed to supporting the design or mounting of a data warehouse with use of a construction support tool. The modification works itself involved by the alteration can be made easy by means of such a tool, but the problems mentioned in items (1), (2) and (3) cannot be solved. In order for the developer to modify the individual alterations to maintain the entire compatibility, it is necessary for the developer to proceed with the modifying works while understanding the entire influence relation.